The games filesize is already over 4(.5?)GB, with each update the filesize is only going to exponentially grow. Potential issues will likely grow as the dev pushed the game/engine.
Why should it grow exponentially? Yes, it will grow further and further (until it's finished obviously) but not necessarily out of the range of the capabilities. There are way larger, well working games. College Kings has 18 or so GB for chapter 1, so four times the size.
The game cant be spontaneous or more functionally sandboxy due to technical limitations and that hampers the devs creative freedom.
Whether it is intended to be more functionally sandboxy is an open question as well. But yeah, technical limitations do exist, and Altos might think of something he would want to do but cannot do but could do in Unity. Maybe he can even already give us a list of content missed for that exact reason.
But as others have pointed out, it would be tremendously time consuming to rework all that stuff even if he is adept at Unity and everything that comes with it. If not, it's not better. And I guess we have a case to guess that not, after all, if you look back, he often mentions things he learned with PlayHome and RenPy he just learned and will make good use of in the future.
I do get your mindset. But in my experience it doesn't work well - there are some devs doing exactly that. They cannot make the game they want to, so they restart with a different engine/animation program/whathaveyou. And it looks great, it works, it feels sensational. Then the next thing comes along. It may not even have to be a new engine, it can only be a new framework you create yourself within the engine to have everything run more smoothly.
And the games are often in their fifth iteration, not advancing one iota in terms of story and content. Of course, early in development you can switch if you see that you are wrong. At a certain point however you need to see that your game won't be perfect, because a perfect game will never be finished. And in my opinion this point has passed.
Never mind that working around limitations (and succeeding) can teach you more about programming and can be more fun for the dev than just having it work, that's not interesting to players of this game. But would I love a freeroam game where I can move around and find, say, Liz, and control her, with changing real time scenes depending on weather and advancement in the storyline, maybe even advancing it a bit? Hell yeah, I would love to. However, I don't think it is feasible for two people in their home who are already a couple of years into a project. Even if money and supporters weren't an issue, some times you just need to move forward instead of polishing out. In my opinion at least.
